Pursuant to an exemplary scenario, during encoding of multimedia frames such as bi-directional coded frames (B-frames), a temporal prediction of each B-frame is performed bi-directionally by referring to one or more reference frames temporally preceding the B-frame and one or more reference frames temporally succeeding the B-frame in an order of capture of the multimedia frames. In the above-described scenario, the temporally succeeding multimedia frames are encoded prior to encoding the B-frame so as to render the temporally succeeding multimedia frames available during the encoding of the B frame. However, multimedia capture devices output the captured multimedia frames as buffers in the order that the multimedia frames are captured, which is different from an order of encoding the multimedia frames.